Professor David Stuart FRS
|Research Area:||Protein Science and Structural Biology|
|Technology Exchange:||Bioinformatics, Computational biology, Crystallography and Protein interaction|
|Scientific Themes:||Protein Science & Structural Biology|
|Keywords:||Viral Proteins, X Ray crystallographic methods, Viruses, virus-receptor interactions, retroviruses and structure to function of genome|
Viruses are attractive targets for study at the molecular level, since they are sufficiently simple that we may hope to achieve a rather complete understanding of their biology. In practice although their genomes are compact they display astonishing diversity, both in structure and function. Our attempts to relate structure to function have benefited from the developments in X-ray crystallographic methods that have brought very complex structures within reach of description in atomic detail. Our targets range from picornaviruses, small ssRNA viruses, which include a number of important animal and human pathogens, to the larger dsRNA viruses. At both ends of this spectrum (from less than 10,000,000 to about 100,000,000 Daltons) we now have representative atomic structures.
Our efforts are particularly focused on virus-receptor interactions and basic puzzles of virus assembly. Our studies here are highly collaborative, with strong links with a number of virologists (P. Mertens and B. Charleston (Pirbright), D. Rowlands (Leeds), P. Roy (London) as well as numerous groups elsewhere in Europe).
Work on cell-surface molecules is largely performed in collaboration with the group of Prof. E.Y. Jones, whose entry describes many of the projects.
We have a particular interest in studying virus evolution and many of these studies are perfoirmed in collaboration with D. Bamford in Helsinki.
Finally, we are studying a number of viral proteins and enzymes which are potential drug targets and/or illuminate how viruses modulate host responses. For example, the immune modulators of pox viruses.
|Professor Dennis Bamford||Institute of Biotechnology and Department of Biosciences||University of Helsinki||Finland|
|Dr Bryan Charleston||Jenner Institute||University of Oxford||United Kingdom|
Picornaviruses are responsible for a range of human and animal diseases, but how their RNA genome is packaged remains poorly understood. A particularly poorly studied group within this family are those that lack the internal coat protein, VP4. Here we report the atomic structure of one such virus, Ljungan virus, the type member of the genus Parechovirus B, which has been linked to diabetes and myocarditis in humans. The 3.78-Å resolution cryo-electron microscopy structure shows remarkable features, including an extended VP1 C terminus, forming a major protuberance on the outer surface of the virus, and a basic motif at the N terminus of VP3, binding to which orders some 12% of the viral genome. This apparently charge-driven RNA attachment suggests that this branch of the picornaviruses uses a different mechanism of genome encapsidation, perhaps explored early in the evolution of picornaviruses. Hide abstract
Polyhedra represent an ancient system used by a number of insect viruses to protect virions during long periods of environmental exposure. We present high resolution crystal structures of polyhedra for seven previously uncharacterised types of cypoviruses, four using ab initio selenomethionine phasing (two of these required over 100 selenomethionine crystals each). Approximately 80% of residues are structurally equivalent between all polyhedrins (pairwise rmsd ⩽ 1.5 Å), whilst pairwise sequence identities, based on structural alignment, are as little as 12%. These structures illustrate the effect of 400 million years of evolution on a system where the crystal lattice is the functionally conserved feature in the face of massive sequence variability. The conservation of crystal contacts is maintained across most of the molecular surface, except for a dispensable virus recognition domain. By spreading the contacts over so much of the protein surface the lattice remains robust in the face of many individual changes. Overall these unusual structural constraints seem to have skewed the molecule's evolution so that surface residues are almost as conserved as the internal residues. Hide abstract
UNLABELLED: Enterovirus 71 (EV71) and coxsackievirus A16 (CVA16) are the primary causes of the epidemics of hand-foot-and-mouth disease (HFMD) that affect more than a million children in China each year and lead to hundreds of deaths. Although there has been progress with vaccines for EV71, the development of a CVA16 vaccine has proved more challenging, and the EV71 vaccine does not give useful cross-protection, despite the capsid proteins of the two viruses sharing about 80% sequence identity. The structural details of the expanded forms of the capsids, which possess nonnative antigenicity, are now well understood, but high resolution information for the native antigenic form of CVA16 has been missing. Here, we remedy this with high resolution X-ray structures of both mature and natural empty CVA16 particles and also of empty recombinant viruslike particles of CVA16 produced in insect cells, a potential vaccine antigen. All three structures are unexpanded native particles and antigenically identical. The recombinant particles have recruited a lipid moiety to stabilize the native antigenic state that is different from the one used in a natural virus infection. As expected, the mature CVA16 virus is similar to EV71; however, structural and immunogenic comparisons highlight differences that may have implications for vaccine production. IMPORTANCE: Hand-foot-and-mouth disease is a serious public health threat to children in Asian-Pacific countries, resulting in millions of cases. EV71 and CVA16 are the two dominant causative agents of the disease that, while usually mild, can cause severe neurological complications, leading to hundreds of deaths. EV71 vaccines do not provide protection against CVA16. A CVA16 vaccine or bivalent EV71/CVA16 vaccine is therefore urgently needed. We report atomic structures for the mature CVA16 virus, a natural empty particle, and a recombinant CVA16 virus-like particle that does not contain the viral genome. All three particles have similar structures and identical antigenicity. The recombinant particles, produced in insect cells (a system suitable for making vaccine antigen), are stabilized by recruiting from the insect cells a small molecule that is different from that used by the virus in a normal infection. We present structural and immunogenic comparisons with EV71 to facilitate structure-based drug design and vaccine development. Hide abstract
Virus capsids are primed for disassembly, yet capsid integrity is key to generating a protective immune response. Foot-and-mouth disease virus (FMDV) capsids comprise identical pentameric protein subunits held together by tenuous noncovalent interactions and are often unstable. Chemically inactivated or recombinant empty capsids, which could form the basis of future vaccines, are even less stable than live virus. Here we devised a computational method to assess the relative stability of protein-protein interfaces and used it to design improved candidate vaccines for two poorly stable, but globally important, serotypes of FMDV: O and SAT2. We used a restrained molecular dynamics strategy to rank mutations predicted to strengthen the pentamer interfaces and applied the results to produce stabilized capsids. Structural analyses and stability assays confirmed the predictions, and vaccinated animals generated improved neutralizing-antibody responses to stabilized particles compared to parental viruses and wild-type capsids. Hide abstract
It remains largely mysterious how the genomes of non-enveloped eukaryotic viruses are transferred across a membrane into the host cell. Picornaviruses are simple models for such viruses, and initiate this uncoating process through particle expansion, which reveals channels through which internal capsid proteins and the viral genome presumably exit the particle, although this has not been clearly seen until now. Here we present the atomic structure of an uncoating intermediate for the major human picornavirus pathogen CAV16, which reveals VP1 partly extruded from the capsid, poised to embed in the host membrane. Together with previous low-resolution results, we are able to propose a detailed hypothesis for the ordered egress of the internal proteins, using two distinct sets of channels through the capsid, and suggest a structural link to the condensed RNA within the particle, which may be involved in triggering RNA release. Hide abstract
Foot-and-mouth disease remains a major plague of livestock and outbreaks are often economically catastrophic. Current inactivated virus vaccines require expensive high containment facilities for their production and maintenance of a cold-chain for their activity. We have addressed both of these major drawbacks. Firstly we have developed methods to efficiently express recombinant empty capsids. Expression constructs aimed at lowering the levels and activity of the viral protease required for the cleavage of the capsid protein precursor were used; this enabled the synthesis of empty A-serotype capsids in eukaryotic cells at levels potentially attractive to industry using both vaccinia virus and baculovirus driven expression. Secondly we have enhanced capsid stability by incorporating a rationally designed mutation, and shown by X-ray crystallography that stabilised and wild-type empty capsids have essentially the same structure as intact virus. Cattle vaccinated with recombinant capsids showed sustained virus neutralisation titres and protection from challenge 34 weeks after immunization. This approach to vaccine antigen production has several potential advantages over current technologies by reducing production costs, eliminating the risk of infectivity and enhancing the temperature stability of the product. Similar strategies that will optimize host cell viability during expression of a foreign toxic gene and/or improve capsid stability could allow the production of safe vaccines for other pathogenic picornaviruses of humans and animals. Hide abstract
Enveloped viruses have developed various adroit mechanisms to invade their host cells. This process requires one or more viral envelope glycoprotein to achieve cell attachment and membrane fusion. Members of the Flaviviridae such as flaviviruses possess only one envelope glycoprotein, E, whereas pestiviruses and hepacivirus encode two glycoproteins, E1 and E2. Although E2 is involved in cell attachment, it has been unclear which protein is responsible for membrane fusion. We report the crystal structures of the homodimeric glycoprotein E2 from the pestivirus bovine viral diarrhea virus 1 (BVDV1) at both neutral and low pH. Unexpectedly, BVDV1 E2 does not have a class II fusion protein fold, and at low pH the N-terminal domain is disordered, similarly to the intermediate postfusion state of E2 from sindbis virus, an alphavirus. Our results suggest that the pestivirus and possibly the hepacivirus fusion machinery are unlike any previously observed. Hide abstract
Enterovirus 71 (EV71) is a major agent of hand, foot and mouth disease in children that can cause severe central nervous system disease and death. No vaccine or antiviral therapy is available. High-resolution structural analysis of the mature virus and natural empty particles shows that the mature virus is structurally similar to other enteroviruses. In contrast, the empty particles are markedly expanded and resemble elusive enterovirus-uncoating intermediates not previously characterized in atomic detail. Hydrophobic pockets in the EV71 capsid are collapsed in this expanded particle, providing a detailed explanation of the mechanism for receptor-binding triggered virus uncoating. These structures provide a model for enterovirus uncoating in which the VP1 GH loop acts as an adaptor-sensor for cellular receptor attachment, converting heterologous inputs to a generic uncoating mechanism, highlighting new opportunities for therapeutic intervention. Hide abstract
Equine rhinitis A virus (ERAV) is closely related to foot-and-mouth disease virus (FMDV), belonging to the genus Aphthovirus of the Picornaviridae. How picornaviruses introduce their RNA genome into the cytoplasm of the host cell to initiate replication is unclear since they have no lipid envelope to facilitate fusion with cellular membranes. It has been thought that the dissociation of the FMDV particle into pentameric subunits at acidic pH is the mechanism for genome release during cell entry, but this raises the problem of how transfer across the endosome membrane of the genome might be facilitated. In contrast, most other picornaviruses form 'altered' particle intermediates (not reported for aphthoviruses) thought to induce membrane pores through which the genome can be transferred. Here we show that ERAV, like FMDV, dissociates into pentamers at mildly acidic pH but demonstrate that dissociation is preceded by the transient formation of empty 80S particles which have released their genome and may represent novel biologically relevant intermediates in the aphthovirus cell entry process. The crystal structures of the native ERAV virus and a low pH form have been determined via highly efficient crystallization and data collection strategies, required due to low virus yields. ERAV is closely similar to FMDV for VP2, VP3 and part of VP4 but VP1 diverges, to give a particle with a pitted surface, as seen in cardioviruses. The low pH particle has internal structure consistent with it representing a pre-dissociation cell entry intermediate. These results suggest a unified mechanism of picornavirus cell entry. Hide abstract
Viral fusion proteins mediate the merger of host and viral membranes during cell entry for all enveloped viruses. Baculovirus glycoprotein gp64 (gp64) is unusual in promoting entry into both insect and mammalian cells and is distinct from established class I and class II fusion proteins. We report the crystal structure of its postfusion form, which explains a number of gp64's biological properties including its cellular promiscuity, identifies the fusion peptides and shows it to be the third representative of a new class (III) of fusion proteins with unexpected structural homology with vesicular stomatitis virus G and herpes simplex virus type 1 gB proteins. We show that domains of class III proteins have counterparts in both class I and II proteins, suggesting that all these viral fusion machines are structurally more related than previously thought. Hide abstract
Recent, primarily structural observations indicate that related viruses, harboring no sequence similarity, infect hosts of different domains of life. One such clade of viruses, defined by common capsid architecture and coat protein fold, is the so-called PRD1-adenovirus lineage. Here we report the structure of the marine lipid-containing bacteriophage PM2 determined by crystallographic analyses of the entire approximately 45 MDa virion and of the outer coat proteins P1 and P2, revealing PM2 to be a primeval member of the PRD1-adenovirus lineage with an icosahedral shell and canonical double beta barrel major coat protein. The view of the lipid bilayer, richly decorated with membrane proteins, constitutes a rare visualization of an in vivo membrane. The viral membrane proteins P3 and P6 are organized into a lattice, suggesting a possible assembly pathway to produce the mature virus. Hide abstract
CD47 is a widely distributed cell-surface protein that acts a marker of self through interactions of myeloid and neural cells. We describe the high-resolution X-ray crystallographic structures of the immunoglobulin superfamily domain of CD47 alone and in complex with the N-terminal ligand-binding domain of signal regulatory protein alpha (SIRPalpha). The unusual and convoluted interacting face of CD47, comprising the N terminus and loops at the end of the domain, intercalates with the corresponding regions in SIRPalpha. We have also determined structures of the N-terminal domains of SIRPbeta, SIRPbeta(2), and SIRPgamma; proteins that are closely related to SIRPalpha but bind CD47 with negligible or reduced affinity. These results explain the specificity of CD47 for the SIRP family of paired receptors in atomic detail. Analysis of SIRPalpha polymorphisms suggests that these, as well as the activating SIRPs, may have evolved to counteract pathogen binding to the inhibitory SIRPalpha receptor. Hide abstract
Nipah and Hendra viruses are emergent paramyxoviruses, causing disease characterized by rapid onset and high mortality rates, resulting in their classification as Biosafety Level 4 pathogens. Their attachment glycoproteins are essential for the recognition of the cell-surface receptors ephrin-B2 (EFNB2) and ephrin-B3 (EFNB3). Here we report crystal structures of both Nipah and Hendra attachment glycoproteins in complex with human EFNB2. In contrast to previously solved paramyxovirus attachment complexes, which are mediated by sialic acid interactions, the Nipah and Hendra complexes are maintained by an extensive protein-protein interface, including a crucial phenylalanine side chain on EFNB2 that fits snugly into a hydrophobic pocket on the viral protein. By analogy with the development of antivirals against sialic acid binding viruses, these results provide a structural template to target antiviral inhibition of protein-protein interactions. Hide abstract
The structure of the membrane-containing bacteriophage PRD1 has been determined by X-ray crystallography at about 4 A resolution. Here we describe the structure and location of proteins P3, P16, P30 and P31. Different structural proteins seem to have specialist roles in controlling virus assembly. The linearly extended P30 appears to nucleate the formation of the icosahedral facets (composed of trimers of the major capsid protein, P3) and acts as a molecular tape-measure, defining the size of the virus and cementing the facets together. Pentamers of P31 form the vertex base, interlocking with subunits of P3 and interacting with the membrane protein P16. The architectural similarities with adenovirus and one of the largest known virus particles PBCV-1 support the notion that the mechanism of assembly of PRD1 is scaleable and applies across the major viral lineage formed by these viruses. Hide abstract
Membranes are essential for selectively controlling the passage of molecules in and out of cells and mediating the response of cells to their environment. Biological membranes and their associated proteins present considerable difficulties for structural analysis. Although enveloped viruses have been imaged at about 9 A resolution by cryo-electron microscopy and image reconstruction, no detailed crystallographic structure of a membrane system has been described. The structure of the bacteriophage PRD1 particle, determined by X-ray crystallography at about 4 A resolution, allows the first detailed analysis of a membrane-containing virus. The architecture of the viral capsid and its implications for virus assembly are presented in the accompanying paper. Here we show that the electron density also reveals the icosahedral lipid bilayer, beneath the protein capsid, enveloping the viral DNA. The viral membrane contains about 26,000 lipid molecules asymmetrically distributed between the membrane leaflets. The inner leaflet is composed predominantly of zwitterionic phosphatidylethanolamine molecules, facilitating a very close interaction with the viral DNA, which we estimate to be packaged to a pressure of about 45 atm, factors that are likely to be important during membrane-mediated DNA translocation into the host cell. In contrast, the outer leaflet is enriched in phosphatidylglycerol and cardiolipin, which show a marked lateral segregation within the icosahedral asymmetric unit. In addition, the lipid headgroups show a surprising degree of order. Hide abstract
Many viruses package their genome into preformed capsids using packaging motors powered by the hydrolysis of ATP. The hexameric ATPase P4 of dsRNA bacteriophage phi12, located at the vertices of the icosahedral capsid, is such a packaging motor. We have captured crystallographic structures of P4 for all the key points along the catalytic pathway, including apo, substrate analog bound, and product bound. Substrate and product binding have been observed as both binary complexes and ternary complexes with divalent cations. These structures reveal large movements of the putative RNA binding loop, which are coupled with nucleotide binding and hydrolysis, indicating how ATP hydrolysis drives RNA translocation through cooperative conformational changes. Two distinct conformations of bound nucleotide triphosphate suggest how hydrolysis is activated by RNA binding. This provides a model for chemomechanical coupling for a prototype of the large family of hexameric helicases and oligonucleotide translocating enzymes. Hide abstract